I am currently attempting to use the heckpoisson command in Stata 15.1. My dependent variable is an annual count of terrorist attacks, data exhibiting both overdispersion and large amounts of zero values. I am receiving this message:

initial: log pseudolikelihood = -<inf> (could not be evaluated)
could not find feasible values
r(491);

My question is, the description for heckpoisson states that this command relaxes the assumption of equidispersion with regular poisson but would the presence of large amounts of zero values in my dependent variable be problematic for this command?

Thanks,
Harrison